Hello.
Maybe after work with FreeBSD I become too spoilt, but I expect that packages from one distribution kit works among themselves well. But that I see:
$ rpm -qa | grep rpm
rpm-libs-4.3.13_nonptl rpm-python-4.3.13_nonptl rpm-4.3.13_nonptl rpm-build-4.3.13_nonptl
$python
import rpm
ImportError /usr/lib/librpm-4.3.so: undefined symbol: rpm_execcon
Thanks.
Dan Shidlovsky wrote:
Maybe after work with FreeBSD I become too spoilt, but I expect that packages from one distribution kit works among themselves well. But that I see:
$ rpm -qa | grep rpm
rpm-libs-4.3.13_nonptl rpm-python-4.3.13_nonptl rpm-4.3.13_nonptl rpm-build-4.3.13_nonptl
$python
import rpm
ImportError /usr/lib/librpm-4.3.so: undefined symbol: rpm_execcon
this works for me just fine.
did you install anything from source / third party repo that might have caused this issue ?
what arch is this ?
can you rpm -V the rpm packages ? is selinux enforcing ?
can you rpm -V the rpm packages ? is selinux enforcing ?
I don't think it's an selinux problem, as the machine I tested on above is my webserver, and is actively using selinux. It let me do it just fine, and I double checked it a couple different ways to be sure.
Jim Perrin wrote:
can you rpm -V the rpm packages ? is selinux enforcing ?
I don't think it's an selinux problem, as the machine I tested on above is my webserver, and is actively using selinux. It let me do it just fine, and I double checked it a couple different ways to be sure.
ok accepted, it will work fine - yum needs this to work and it does work fine either with or without selinux, but its easy for people new to selinux to write policy that can kick their own butt - and in most cases, blame everyone/ everything else.
there isnt enough info in there to predict what is going on - but a driveby python or rpm hackerage seems to be to blame. Maybe thats how they manage packages on FreeBSD :)
- KB
Maybe after work with FreeBSD I become too spoilt, but I expect that packages from one distribution kit works among themselves well. But that I see:
Nothing like biting wit to start the morning.
$ rpm -qa | grep rpm
rpm-libs-4.3.13_nonptl rpm-python-4.3.13_nonptl rpm-4.3.13_nonptl rpm-build-4.3.13_nonptl
Seems I've got the same versions you do. Couple additional packages, but nothing major. [jperrin@www ~]$ rpm -qa | grep rpm rpm-libs-4.3.3-13_nonptl rpm-build-4.3.3-13_nonptl rpm-python-4.3.3-13_nonptl rpmdb-CentOS-4.3-0.20060314 redhat-rpm-config-8.0.32.1-1 rpm-4.3.3-13_nonptl rpm-devel-4.3.3-13_nonptl
$python
import rpm
ImportError /usr/lib/librpm-4.3.so: undefined symbol: rpm_execcon
Yep. That's an error alright. However it works fine here.
[jperrin@www ~]$ python Python 2.3.4 (#1, Mar 10 2006, 06:12:09) [GCC 3.4.5 20051201 (Red Hat 3.4.5-2)] on linux2 Type "help", "copyright", "credits" or "license" for more information.
import rpm
Now call me suspicious, but your python does not identify itself, and you don't seem to list the python version. Did you upgrade python?
Dan Shidlovsky wrote:
Now call me suspicious, but your python does not identify itself, and you don't seem to list the python version. Did you upgrade python?
No.
Ermm.
| $ rpm -qa | grep rpm | | rpm-libs-4.3.13_nonptl | rpm-python-4.3.13_nonptl | rpm-4.3.13_nonptl | rpm-build-4.3.13_nonptl
Was that cut and paste? There's no such thing as 4.3.13:
rpm-libs-4.3.3-13_nonptl rpm-build-4.3.3-13_nonptl rpm-python-4.3.3-13_nonptl rpm-4.3.3-13_nonptl
$ python Python 2.3.4 (#1 Feb 22 2005, 04:09:37) [GCC 3.4.3 20041212 (Red Hat 3.4.3.-9.EL4) on linux2 Type "help" ...
What does 'rpm -q python' say?
Ralph
Dan Shidlovsky wrote:
Was that cut and paste? There's no such thing as 4.3.13:
Oh, yes. You are right. There is 4.3.3-13_nonptl.
What does 'rpm -q python' say?
python-2.3.4-14.1
Then you have a second python on that machine which comes first in $PATH.
Ralph
Ralph Angenendt wrote:
Dan Shidlovsky wrote:
Was that cut and paste? There's no such thing as 4.3.13:
Oh, yes. You are right. There is 4.3.3-13_nonptl.
What does 'rpm -q python' say?
python-2.3.4-14.1
Then you have a second python on that machine which comes first in $PATH.
Hmmm. Maybe not. Looks like some other breakage.
Ralph
Dan Shidlovsky wrote:
Was that cut and paste? There's no such thing as 4.3.13:
Oh, yes. You are right. There is 4.3.3-13_nonptl.
What does 'rpm -q python' say?
python-2.3.4-14.1
so where did the other python ( the one from 4.0 ) come from ?
and you've not answered any of my earlier questions - what arch is this ?
- K
On 6/21/06, Dan Shidlovsky dmitrikozhevin@gmail.com wrote:
Erm, that is most definitely *not* a CentOS kernel...
I'v got this system with virtual hosting.
Out of sheer morbid curiosity, what version of CentOS does this claim to be (cat /etc/redhat-release)? I don't recall this being mentioned in the threads, however the software versions would seem to indicate that it's based on CentOS 4, however with a 2.4 based kernel, I can only envision bad things coming from this. CentOS 4 was built around the 2.6 kernel, selinux, and udev support... all of which I would think a 2.4 kernel would throw a hissy fit over.
To throw back to your original post, CentOS does work fine with the distributed software, as our primary update system (yum) relies on exactly this call to kick things off. You however seem to be using some flavor of retarded mutant frankenstein version, and I would hold your VPS provider entirely responsible for the train wreck you have there.
Can't install emacs, at all. Terrible. :(
That's a feature, not a flaw. /vim user for life //flame suit on!
Gotta have a laugh now and then.. sorry.
On Wed, 21 Jun 2006, Dan Shidlovsky wrote:
Erm, that is most definitely *not* a CentOS kernel...
I'v got this system with virtual hosting. Can't install emacs, at all. Terrible. :(
Errmm , this doesnt seem like the correct CentOS components to install, even for virtual hosting.
Perhapos you could let us know who you got the system from, so that we can contact them and discuss how they can make things work properly, and avoid issues in the future.
Thanks
Regards Lance
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Dan Shidlovsky wrote:
and you've not answered any of my earlier questions - what arch is this ?
AMD Opteron(tm) Processor 248 Linux 2.4.20-021stab028.24.777-enterprise
humm a 2.4 kernel on CentOS4 is going to cause much issues.... exactly what did you do to end up like this ? and more importantly 'why ?'
that being beside the point for now, how do you intend to keep the machine updated for bug and security fix's if there is no working package manager ?
if you really wanted to be persistent, you could just manually download the emacs rpm's from http://mirror.centos.org/centos/4/os/i386/CentOS/RPMS/ and manually rpm -i them in. but I'd still try and get a proper fix for the selinux issue and get a package manager working
- KB
uklinux.net - The ISP of choice for the discerning Linux user.
On Wed, 21 Jun 2006, Dan Shidlovsky wrote:
Was that cut and paste? There's no such thing as 4.3.13:
Oh, yes. You are right. There is 4.3.3-13_nonptl.
What does 'rpm -q python' say?
python-2.3.4-14.1
What does rpm -q libselinux say
and can you see /lib/libselinux.so.1 ??
cos that is where rpm_execcon lives ...
Regards Lance
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Ralph Angenendt wrote:
Dan Shidlovsky wrote:
$ python Python 2.3.4 (#1 Feb 22 2005, 04:09:37) [GCC 3.4.3 20041212 (Red Hat 3.4.3.-9.EL4) on linux2 Type "help" ...
What does 'rpm -q python' say?
I ask that, because you have the rpm version of CentOS 4.3 in there, but seem to have a python version from CentOS 4.0 - the build date is quite old, a python matching your version of RPM looks should have a build date in 2006.
2.3.4-14.1 was in stock 4.3, there was an update to 2.3.4-14.2 later on.
Ralph
On Wed, 2006-06-21 at 08:16 -0400, Jim Perrin wrote:
Maybe after work with FreeBSD I become too spoilt, but I expect that packages from one distribution kit works among themselves well.
Glad to see you too this fine A.M. :-=)
<snip>
Seems I've got the same versions you do. Couple additional packages, but nothing major. [jperrin@www ~]$ rpm -qa | grep rpm rpm-libs-4.3.3-13_nonptl rpm-build-4.3.3-13_nonptl rpm-python-4.3.3-13_nonptl rpmdb-CentOS-4.3-0.20060314 redhat-rpm-config-8.0.32.1-1 rpm-4.3.3-13_nonptl rpm-devel-4.3.3-13_nonptl
[root@wlmlfs08 ~]# rpm -qa rpm* rpm-python-4.3.3-13_nonptl.i386 rpm-4.3.3-11_nonptl.i386 rpm-build-4.3.3-11_nonptl.i386 rpmdb-CentOS-4.2-0.20051011.i386 rpmdb-CentOS-4.3-0.20060314.i386 rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386 rpm-build-4.3.3-13_nonptl.i386 rpm-4.3.3-13_nonptl.i386 rpm-devel-4.3.3-11_nonptl.i386 rpmforge-release-0.3.4-1.el4.rf.i386 rpm-python-4.3.3-11_nonptl.i386 rpm-devel-4.3.3-13_nonptl.i386 [root@wlmlfs08 ~]#
$python
import rpm
ImportError /usr/lib/librpm-4.3.so: undefined symbol: rpm_execcon
Yep. That's an error alright. However it works fine here.
[jperrin@www ~]$ python Python 2.3.4 (#1, Mar 10 2006, 06:12:09) [GCC 3.4.5 20051201 (Red Hat 3.4.5-2)] on linux2 Type "help", "copyright", "credits" or "license" for more information.
import rpm
Works fine here too.
[root@wlmlfs08 ~]# python Python 2.3.4 (#1, Mar 10 2006, 06:12:09) [GCC 3.4.5 20051201 (Red Hat 3.4.5-2)] on linux2 Type "help", "copyright", "credits" or "license" for more information.
import rpm
[root@wlmlfs08 ~]#
Now call me suspicious, but your python does not identify itself, and you don't seem to list the python version. Did you upgrade python?
<sarcasm> You imply that it might be a "wetware" problem rather than non-BSD (i.e CentOS) incompatibility? I knew you were a cranky old fart! ;-) </sarcasm>
William L. Maltby wrote:
rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386
Well, that also looks completely broken to me.
Ralph
On Wed, 2006-06-21 at 14:40 +0200, Ralph Angenendt wrote:
William L. Maltby wrote:
rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386
Well, that also looks completely broken to me.
Thanks Ralph. That got my eyes focused on the right problem. I'm sure you'll see my other post, so I won't repeat it all here.
<snip>
William L. Maltby wrote:
[root@wlmlfs08 ~]# rpm -qa rpm* rpm-python-4.3.3-13_nonptl.i386 rpm-4.3.3-11_nonptl.i386 rpm-build-4.3.3-11_nonptl.i386 rpmdb-CentOS-4.2-0.20051011.i386 rpmdb-CentOS-4.3-0.20060314.i386 rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386 rpm-build-4.3.3-13_nonptl.i386 rpm-4.3.3-13_nonptl.i386 rpm-devel-4.3.3-11_nonptl.i386 rpmforge-release-0.3.4-1.el4.rf.i386 rpm-python-4.3.3-11_nonptl.i386 rpm-devel-4.3.3-13_nonptl.i386 [root@wlmlfs08 ~]#
you obviously have a fubar'd rpmdb and/or a broken update/upgrade that you never fixed!
did you just never realise there were duplicates there ?
On Wed, 2006-06-21 at 13:40 +0100, Karanbir Singh wrote:
William L. Maltby wrote:
[root@wlmlfs08 ~]# rpm -qa rpm* rpm-python-4.3.3-13_nonptl.i386 rpm-4.3.3-11_nonptl.i386 rpm-build-4.3.3-11_nonptl.i386 rpmdb-CentOS-4.2-0.20051011.i386 rpmdb-CentOS-4.3-0.20060314.i386 rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386 rpm-build-4.3.3-13_nonptl.i386 rpm-4.3.3-13_nonptl.i386 rpm-devel-4.3.3-11_nonptl.i386 rpmforge-release-0.3.4-1.el4.rf.i386 rpm-python-4.3.3-11_nonptl.i386 rpm-devel-4.3.3-13_nonptl.i386 [root@wlmlfs08 ~]#
you obviously have a fubar'd rpmdb and/or a broken update/upgrade that you never fixed!
did you just never realise there were duplicates there ?
Never looked before. Remember, this is new to me. Anyway, when I saw that, I thought of two possibilities and decided to investigate. One is the possibility you mention. The other is the possibility that answers a question I posted elsewhere: is "rpm -qa <pkg>*" reliable.
I posted as I did only so that the WFM answer Jim posted had some immediate support.
But thanks for pointing it out. Being new, I could have overlooked it entirely. I'm hoping that its the "... rpm*" that is the cause rather than the assumption that it's obviously fubar'd.
On Wed, 2006-06-21 at 08:48 -0400, William L. Maltby wrote:
On Wed, 2006-06-21 at 13:40 +0100, Karanbir Singh wrote:
William L. Maltby wrote:
[root@wlmlfs08 ~]# rpm -qa rpm* rpm-python-4.3.3-13_nonptl.i386 rpm-4.3.3-11_nonptl.i386 rpm-build-4.3.3-11_nonptl.i386 rpmdb-CentOS-4.2-0.20051011.i386 rpmdb-CentOS-4.3-0.20060314.i386 rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386 rpm-build-4.3.3-13_nonptl.i386 rpm-4.3.3-13_nonptl.i386 rpm-devel-4.3.3-11_nonptl.i386 rpmforge-release-0.3.4-1.el4.rf.i386 rpm-python-4.3.3-11_nonptl.i386 rpm-devel-4.3.3-13_nonptl.i386 [root@wlmlfs08 ~]#
you obviously have a fubar'd rpmdb and/or a broken update/upgrade that you never fixed!
I think I'm seeing the light here. Combined with what Ralph said, which was most helpful, I apparently am not supposed to have multiple versions laying around? Unlike the kernel?
did you just never realise there were duplicates there ?
When you said "duplicates", I was thinking along the lines of two "4.3.3-11", etc. But I see that is not the case here.
[root@wlmlfs08 ~]# # Save file for diff w/"rpm*" vers for other thread [root@wlmlfs08 ~]# rpm -qa|grep rpm|sort >/tmp/tmp2;cat /tmp/tmp2 redhat-rpm-config-8.0.32.1-1.noarch rpm-4.3.3-11_nonptl.i386 rpm-4.3.3-13_nonptl.i386 rpm-build-4.3.3-11_nonptl.i386 rpm-build-4.3.3-13_nonptl.i386 rpmdb-CentOS-4.2-0.20051011.i386 rpmdb-CentOS-4.3-0.20060314.i386 rpm-devel-4.3.3-11_nonptl.i386 rpm-devel-4.3.3-13_nonptl.i386 rpmforge-release-0.3.4-1.el4.rf.i386 rpm-libs-4.3.3-11_nonptl.i386 rpm-libs-4.3.3-13_nonptl.i386 rpm-python-4.3.3-11_nonptl.i386 rpm-python-4.3.3-13_nonptl.i386
So I now presume, based on the snippet Ralph provided that the breakage is not *duplicates*, but multiple versions installed at the same time (if we could believe my rpm db)? Is that correct?
<snip>
TIA
William L. Maltby wrote:
So I now presume, based on the snippet Ralph provided that the breakage is not *duplicates*, but multiple versions installed at the same time (if we could believe my rpm db)? Is that correct?
i should have been clearer initially, its duplicates as in multiple copies of the same rpm but with different versions. You only want to have 1 version of 1 rpm installed per Arch
On Wed, 2006-06-21 at 14:51 +0100, Karanbir Singh wrote:
William L. Maltby wrote:
So I now presume, based on the snippet Ralph provided that the breakage is not *duplicates*, but multiple versions installed at the same time (if we could believe my rpm db)? Is that correct?
i should have been clearer initially, its duplicates as in multiple copies of the same rpm but with different versions. You only want to have 1 version of 1 rpm installed per Arch
OK. Thanks. When I first started CentOS and had almost *zero* yum and rpm background, I had some failures during updates. Learned about clearing the yum cache and headers (IIRC?) from you folks on the list, but had not yet read all the yum and rpm man pages and other docs. So (out of fear or haste intelligence? ;-) never looked into the rpm status. Things were working, I was busy reading/learning and I was "happy".
Now I'll "rpm -qa" it and see what I can find. Must be other stuff. Anyway, thanks for the heads up.
On 21/06/06, William L. Maltby BillsCentOS@triad.rr.com wrote:
On Wed, 2006-06-21 at 14:51 +0100, Karanbir Singh wrote:
William L. Maltby wrote:
So I now presume, based on the snippet Ralph provided that the breakage is not *duplicates*, but multiple versions installed at the same time (if we could believe my rpm db)? Is that correct?
i should have been clearer initially, its duplicates as in multiple copies of the same rpm but with different versions. You only want to have 1 version of 1 rpm installed per Arch
OK. Thanks. When I first started CentOS and had almost *zero* yum and rpm background, I had some failures during updates. Learned about clearing the yum cache and headers (IIRC?) from you folks on the list, but had not yet read all the yum and rpm man pages and other docs. So (out of fear or haste intelligence? ;-) never looked into the rpm status. Things were working, I was busy reading/learning and I was "happy".
Now I'll "rpm -qa" it and see what I can find. Must be other stuff. Anyway, thanks for the heads up.
You might want to go have a look back though the archives for the cause of these problems.
http://lists.centos.org/pipermail/centos/2006-March/062568.html
This will give you a list of duplicated packages...
# rpm -qa --qf "%{NAME}.%{arch}\n" | sort | uniq -d
Will.
On Wed, 2006-06-21 at 15:19 +0100, Will McDonald wrote:
On 21/06/06, William L. Maltby BillsCentOS@triad.rr.com wrote:
On Wed, 2006-06-21 at 14:51 +0100, Karanbir Singh wrote:
William L. Maltby wrote:
<snip>
Now I'll "rpm -qa" it and see what I can find. Must be other stuff. Anyway, thanks for the heads up.
You might want to go have a look back though the archives for the cause of these problems.
http://lists.centos.org/pipermail/centos/2006-March/062568.html
This will give you a list of duplicated packages...
# rpm -qa --qf "%{NAME}.%{arch}\n" | sort | uniq -d
Thanks Will! Why is it so easy to recall reading all that once someone points you to it but I'm a total blank about it (sometimes) before being "reoriented". OTOH, I remember with amazing clarity most authored by myself! :-)
Will.
<snip sig stuff>
On 21/06/06, William L. Maltby BillsCentOS@triad.rr.com wrote:
You might want to go have a look back though the archives for the cause of these problems.
http://lists.centos.org/pipermail/centos/2006-March/062568.html
This will give you a list of duplicated packages...
# rpm -qa --qf "%{NAME}.%{arch}\n" | sort | uniq -d
Thanks Will! Why is it so easy to recall reading all that once someone points you to it but I'm a total blank about it (sometimes) before being "reoriented". OTOH, I remember with amazing clarity most authored by myself! :-)
Good question. I probably remembered that thread because I suffered the problem and contributed a little (even though it was just to point out the previous thread). :)
I suppose we skim threads and only really notice those that impact us directly. I know I rarely read threads on X config for example, so if I ever encounter X problems I'll be semi-clueless and will no doubt ask questions that've been covered before.
Will.
-
Dan Shidlovsky -
Jim Perrin -
Joshua Baker-LePain -
Karanbir Singh -
Lance Davis -
Lance Davis -
Ralph Angenendt -
Will McDonald -
William L. Maltby